@Pecos an approach I am evaluating is creating an encloser for the packs made of thermal conductive filament. This will conduct the heat out of the cells, the only challenge is that we need to be careful since the filament could be conductive as well, We need to make sure is well insulated on that cathode.
Filament reference: https://amzn.to/46gMj6W
I think youâd be better off potting the battery in a thermally conductive epoxy and then attaching a heatsink to that
Can you point me in that Epoxy?? I have never use one thermally conductive.
The reason I know this stuff works is covered by an NDA I signed.
I hear you !! Thanks !
Apparently, once one adds about 30-40% of copper powder to casting epoxy, by volume, the thermal conductivity goes way up.
If added weight is not a concern, casting P groups together seems viable.
Got to be better than a dead air space in transferring cell heat to the enclosure.
Definitely.
AnythingâŚliterally ANYTHINGâŚwe might use will have higher thermal conductivity than air. 
I wonder what percentage of copper by weight could be added to the epoxy before the conductivity becomes a problem.
Maybe filling it with ceramic beads instead would mitigate that issue while still gaining some thermal conductivity
AI says:
Materials like boron nitride (BN), aluminum nitride (AlN), and silicon nitride (SiN) offer high thermal conductivity and electrical insulation.
I rebuilt my 3s2p Laptop battery using EVE 35V cells last night using 0.15mm pure nickel.
Gear 65 of 999 was adequate. Single pulse. 0.05 ms preheating
The nickel would tear, but if an esk8 battery Iâd want bigger dots of nickel left on the cell.
My tear off strength finger dynometer has been recalibrated with all my other testing with raw copper and copper steel sandwich though.
Perhaps 0.2mm copper by itself is not too weak. The steel sandwich adds so much tear off strength that it corrupts tear off tests when it is not used.
But this leaves the question how strong do the welds need to be? For a no vibration , no physical stress laptop battery, 4 weak welds/dots, by esk8 standards, seems more than adequate.
This battery with the Wintonic cells used 0.1 nickel plated steel, choked to 6mm wide in parts.
I had unnecessarily added some fishpaper on top of parallel and series strips before folding the series connections, but had to remove it as they added too much length end to end, to fit in the plastic casing.
Just adding the fishpaper rings to the EVE35v, and the 0.3mm worth of nickel between the 2p groups had it barely fit.
The laptop had no issues waking the BMS and charging it to what it said was full.
I removed power cord and got an hour browsing old photos, and it said 67% and 1.57 hours remaining, but no doubt it needs to learn it has a lot more capacity available, with 7000mah instead of the degraded 4400.
The laptop power supply is 19.5vdc and 80 watts, so there has to be some buck and boost conversions going on within for an 11.1v nominal battery to be able to power it.
I later did some experiments with copper and nickel steel buttontops. I have 4 devices which require buttontops. These button tops are 0.15mm, and they throw huge sparks, and have steel stick to the probes, which will throw off each subsequent weld if not filed back to copper.
I was able to get copper welded to button top exterior using gear 004 button upside down, and welded to the cell with gear 085, but one weld was good, and the other would often blow through the button and 0.1mm copper.
I decided the desire for copper clad buttontops could only slightly benefit one of the 3 devices I have which require a buttontop, that the consistency was too bad, and not worth the risk.
I welded a button to an EVE 30P, using gear 004. The flashlight was happy enough with a 35v. The 30p should have less sag and run cooler even if it canât run it as long on lesser settings.
Perhaps my lux meter app might show some difference.
Probably remove 7 Sunpower 2500mah supposed 25amp CDR cells from my cordless vaccuum battery, and replace with Ampace JP30 later today. I have to be too frugal with the power settings to finish the house with the sunpowers, and when vacuuming rugs, why use 70% power?
If ur going to pot a complete battery thatâs a lot of resin needed and you should use like a 60-65d durometer polyurethane as an epoxy is so rigid it gives no room for the slight expansion of cells that happens and can crack. Doing a very thin layer, or conformal coating, seems more appropriate than potting unless itâs going to be subject to extreme abuse.
Iâm guessing but think adding powdered filler will further harden it.
On Monday, a 2 meter roll of 200mm wide 0.07mm Stainless steel that was advertised as 304 stainless arrived.
It is shinier, more mirror like, than the 0.1mm stainless Nelvick sent me, and is also slightly more magnetic.
Thus, I suspect it is not really 304.
I tested it on 0.2 copper, no flux, same settings, same cell as when I checked for Hummieee, and I could tell no difference.
I also tried to weld dual stacked 0.2mm at gear 940 with flux, and I could not pull it off with my fingers, but it popped off easily with pliers. Tried again with 0.2mm and 0.15mm and was having issues getting the 0.15mm stick to the 0.2, but the 0.2 was stuck to the cell quite well. I think the oxidized copper was the cause, rather than the 0.7mm stainless.
But I am pretty much dissatisfied with the 0.07mm stainless experiment. It does not seem to weld any hotter, allowing for lesser power to achieve the same weld strength, and it does tear easier
I did use it to install Ampace JP30âs into my 7s1p cordless vaccuum battery, on top of 0.2mm copper.
Gear 550, doublepulse 0.10ms preheat interval 02ms.
My digital calipers measured 0.17mm consistently on the nickel plated steel series connections of the original build, on the Sunpower 2500mah/25 amp cells.
The vaccuum battery before disassembly and rebuild, had been on the charger all day, said 100%. The cells were all 4.04v.
The Ampace JP30 cells all read 3.47v when I welded them in.
When I plugged it into the charger it said 99%. I turned the vacuum on and cranked the power to high and could easily tell it was spinning faster than it ever has.
Iâm gonna put 5 2500mah Sunpower cells in an old Ryobi 5s1p battery which used to have LG 1500mah 22amp CDR cells.
I have a 5s1p ridgid battery, that I believe uses Lishen 2000mah 30 amp cells. I cant see their labels without disassembly. The series connections look like 0.15 or 0.2mm nickel plated copper, but it is weird in that only the spot welds touch the cells. There is an ~ 0.15mm gap between strips and the cell can. Probably gonna put JP30âs in here at some point
Just vacuumed the whole house at top speed nearly the whole time, and started going back over areas until the vacuum shut down on its own. The SOC% display stayed at 1% for a few minutes before it did shut down.
It used to shut down when reading 12-18%, if I had it running on higher speeds at that SOC, and if I was not close to being done, had to try and finish at near minimum speed.
The spinning brush was obviously spinning way faster too. Practically pulling the vacuum forward.
This is such a vast improvement in performance of this vacuum, I felt I had to share.
These Ampace JP30âs are damn impressive cells.
I imagine any board that previously used Samsung 30Q , would benefit greatly by rebuilding using these cells.
Replacing those 0.17mm nickel plated steel series connections with wider 0.2 copper played a big part too I am sure.
Still impressed with the cordless Vacuumâs performance, days later with the Ampace JP30âs.
Thank you Nelvick!
Long After Iâve run out of things to vacuum, it still rests on 1% for a minute or three before actually shutting off automatically, all rugs are vacuumed at 100% speed the whole time.
75% on the tile floors, 100% speed on the rugs, and screw all those lower battery conservation speeds that were once necessary in order to finish. it is so much better now.
Took 5 of the 7 Sunpower 2500/ 25 amp CDR cells from vacuum battery, and put them in 2012 era Ryobi 18v 5s1p battery that was originally a 1.5AH battery.
Used 0.2mm copper capped with 0.07mm stainless no flux, gear 550 double pulse, 0.10ms preheat. Didnât take pics but nothing surprising.
Iâve only apretty weak centrifugal style leaf blower to use this battery in, and a still functioning stock Ryobi 4Ah battery has been OK. Nice to have two batteries for it again, without having to use the Ridgid to Ryobi adapter. The Sunower celled 2.5Ah 5s1p battery seems to last at least 75% as long as the original 5s2p 4.0ah, So I consider it a success.
The old Ryobi battery was built poorly in comparison to the 2020 era Ridgid 2.0. The overlength balance wires have some press fit immobilizers, but they can still vibrate.
Iâve held off on the Ridgid 5S1P JP30 upgrade for now.
This here is a single 18650 cell Ryobi Screwdriverâs battery. I have no Idea when my Dad bought this tool. I have been using it for light stuff, but it is a pretty weak tool, and I donât know how much of that is just old battery sag, vs weak tool.
I had to use a Japanese flush cut pull saw to open it up carefully.
It is a 10 amp CDR, 1200mah Sanyo cell.
UR18650SA R2111
Appears to be a 15 amp fuse between the 0.3mm tab on Anode and Upto the PCB.
I thought the spot welded tab to crimped stranded wire connection to be interesting. Think it is Nickel plated steel, 0.3mm, as it seems more magnetic than pure nickel. I donât see pure nickel crimping very well, but honestly donât know, yet. The fishpaper ring is very thick.
My 2021 era Ridgid 5s1p 2.0AH battery looks to be welded with same welder, and only the 4 spot welds/ dots make contact with the cell . The rest of the tab is elevated above, nearly perfectly evenly, with a 0.15mm gap.
Wondering how Overkill a JP30 will be for this tool. Even If it cannot take advantage of the low voltage drop, an old 1200 mag cell vs a tabless 3000Mah should be a no brainer, but not if the magic smoke is released as it was never designed for such a beast of a cell that laughs at a 10 amp load.
Wonder if that 15 amp fuse and the rest of the PCB, would still be adequate with a monster beast JP 30 in there, and if the JP 30 should be better saved for another device that can utilize its impressive capabilities.
I have less capable new and newish cells I could use instead.
Tenpower 32HE or 26HE
Eve 35V
DMEGC 30P
Sunpower INR18650-2500.
Opinions?
I donât think that adding the JP30 will in any way damage the tool as long as everything else stays the same.
although if it is fitted with a 15a fuse, then thereâs no way the tool would be asking for any more than that, so perhaps a lower specced cell would be better suited
Thereâs a lot to read here and not sure what thickness is best. I got .2mm copper and .1mm 304 stainless on route. Good enough? Will see if my âgearsâ on the Amazon welder have whatâs needed
.2mm copper with .1mm 304 is a perfect combination even 0.3mm copper. Depending the welder you have.
Should be here in a couple days. Will post results w this welder
https://www.amazon.com/gp/aw/d/B0DCBVB8XF?psc=1&ref=ppx_pop_mob_b_asin_title
$109 for what has shown to work really well so far.
What material thickness were you welding, and using what setting / % of max power on that welder?
Glad youâre trying this with something other than an AwithZ welder. I would love to try and work copper into an upcoming 18s4p build but wasnât sure if it was feasible with a malectrics and car battery.
@DIY500AMP.COM , any plans to stock 0.2mm x 30mm copper and 0.1mm x 30mm stainless? Iâm great at procrastinating so all good if it takes a couple few weeks.
